Trauma and overuse of joints can lead to painful defects in the articular cartilage and underlying bone, which adversely affect the mechanical and biological function of the entire joint. Over time this early-stage damage can spread, leading to end-stage osteoarthritis (OA). While treatments for osteochondral defects have been developed to relieve pain and delay the spread of damage, current solutions are perceived as unreliable - requiring early surgical revision. To better treat these defects, we have developed an innovative non- degradable, off-the-shelf device that will provide immediate structural integrity to the defec site for the duration of implantation, while also integrating with the host tissue. The device is non-cell based, non-biodegradable, synthetic and porous, and as such represents a significant shift in the current paradigm for the treatment of chondral and osteochondral defects. The implant consists of a solid cylindrical poly(vinyl) alcohol (PVA) core (to resist joint load) concentrically surrounded by a porous PVA outer rim (press-fit with surrounding tissue and designed to integrate with cartilage), with the solid core attached to a porous metal base (for initial fixation and integration with bone). The device is arthroscopically implanted into the defet in a dehydrated form, which then rehydrates in situ to form a strong interface with the host tissue, thus enabling immediate weight bearing. While a series of in vitro and in vivo animal models have demonstrated that the device can integrate with host tissue and mechanically function much in the way of the native tissue, we encountered several instances of mechanical failure of the device when the PVA disassociated from the metal base. The objective of this study is to modify the PVA-metal interface to prevent failures, while maintaining the ability of th device to mechanically function in a loaded joint. To this end we have used finite element models to direct modifications to: (i) the macroscopic geometry of the PVA-metal interface, and (ii) the stiffness of the PVA-metal interface. Our goal is to determine which combination of changes in macroscopic interlock, and PVA stiffness (15 groups in total) are robust enough to withstand sustained weight bearing. First, we will optimize the shear and tensile strengths of the PVA-metal interface to avoid mechanical failure after device rehydration (Specific Aim 1). Secondly, we will subject the device to repetitive axial and shear forces using a custom rolling-sliding device to simulate the forces applied to the device in situ (Specific Aim 2). Optimized design criteria will be based on the maximum shear and tensile stresses at the PVA-metal interface calculated in the FE model (Aim 1), and characterization of the fatigue properties of the device (Aim 2). Through this suite of tests, we will identify the device design with the highest shear and tensile safety factor that is able to restore load distribution across the joint over extended cycles of loading. The optimal design from this study will be subsequently used for re- implantation into our previously established in vivo horse model.